Optimized deep learning model for diagnosing tonsil and adenoid hypertrophy through X-rays

ObjectiveTo explore the application of a deep learning model based on lateral nasopharyngeal X-rays in diagnosing tonsillar and adenoid hypertrophy.MethodsA retrospective study was conducted using DICOM images of lateral nasopharyngeal X-rays from pediatric outpatients aged 2-12 at our hospital from...

Full description

Saved in:
Bibliographic Details
Main Authors: Zhiqing Wu, Ran Zhuo, Yali Yang, Xiaobo Liu, Bin Wu, Jian Wang
Format: Article
Language:English
Published: Frontiers Media S.A. 2025-03-01
Series:Frontiers in Oncology
Subjects:
Online Access:https://www.frontiersin.org/articles/10.3389/fonc.2025.1508525/full
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850251027815596032
author Zhiqing Wu
Ran Zhuo
Yali Yang
Xiaobo Liu
Bin Wu
Jian Wang
author_facet Zhiqing Wu
Ran Zhuo
Yali Yang
Xiaobo Liu
Bin Wu
Jian Wang
author_sort Zhiqing Wu
collection DOAJ
description ObjectiveTo explore the application of a deep learning model based on lateral nasopharyngeal X-rays in diagnosing tonsillar and adenoid hypertrophy.MethodsA retrospective study was conducted using DICOM images of lateral nasopharyngeal X-rays from pediatric outpatients aged 2-12 at our hospital from July 2014 to July 2024. The study included patients exhibiting varying degrees of respiratory obstruction symptoms (disease group). Initially, 1006 images were collected, but after excluding low-quality images and standardizing the imaging phase, 819 images remained. These images were divided into training and validation sets in an 8:2 ratio. The independent test set is consisted of 484 images. We delineated the target areas for tonsils and adenoids and used a YOLOv8n-based model for object detection and use various convolutional neural network models to classify the cropped images, assessing the severity of tonsillar and adenoid hypertrophy. We compared the performance of these models on the training and validation sets using metrics such as ROC-AUC, accuracy, precision, recall, and F1 score.ResultsThe combined model, incorporating YOLOv8 for object detection and secondary classification, demonstrated excellent performance in diagnosing tonsillar and adenoid hypertrophy, significantly improving diagnostic accuracy and consistency. The ResNet18 model, due to its lightweight nature and minimal computational resource requirements, performed exceptionally well in the YOLOv8-ResNet fusion model for detecting and classifying tonsils and adenoids, making it our preferred model.ConclusionThe deep learning model combining YOLOv8n and ResNet18 based on lateral nasopharyngeal X-rays demonstrates significant advantages in diagnosing pediatric tonsillar and adenoid hypertrophy.
format Article
id doaj-art-5c6103e3342d4374bcb86b5250c4e2b2
institution OA Journals
issn 2234-943X
language English
publishDate 2025-03-01
publisher Frontiers Media S.A.
record_format Article
series Frontiers in Oncology
spelling doaj-art-5c6103e3342d4374bcb86b5250c4e2b22025-08-20T01:58:00ZengFrontiers Media S.A.Frontiers in Oncology2234-943X2025-03-011510.3389/fonc.2025.15085251508525Optimized deep learning model for diagnosing tonsil and adenoid hypertrophy through X-raysZhiqing Wu0Ran Zhuo1Yali Yang2Xiaobo Liu3Bin Wu4Jian Wang5Department of Pediatric Surgery, Children's Hospital of Soochow University, Suzhou, Jiangsu, ChinaDepartment of Pediatric Surgery, Children's Hospital of Soochow University, Suzhou, Jiangsu, ChinaIntensive Care Unit, Children's Hospital of Soochow University, Suzhou, Jiangsu, ChinaDepartment of Pediatric Surgery, Children's Hospital of Soochow University, Suzhou, Jiangsu, ChinaDepartment of Pediatric Surgery, Children's Hospital of Soochow University, Suzhou, Jiangsu, ChinaDepartment of Pediatric Surgery, Children's Hospital of Soochow University, Suzhou, Jiangsu, ChinaObjectiveTo explore the application of a deep learning model based on lateral nasopharyngeal X-rays in diagnosing tonsillar and adenoid hypertrophy.MethodsA retrospective study was conducted using DICOM images of lateral nasopharyngeal X-rays from pediatric outpatients aged 2-12 at our hospital from July 2014 to July 2024. The study included patients exhibiting varying degrees of respiratory obstruction symptoms (disease group). Initially, 1006 images were collected, but after excluding low-quality images and standardizing the imaging phase, 819 images remained. These images were divided into training and validation sets in an 8:2 ratio. The independent test set is consisted of 484 images. We delineated the target areas for tonsils and adenoids and used a YOLOv8n-based model for object detection and use various convolutional neural network models to classify the cropped images, assessing the severity of tonsillar and adenoid hypertrophy. We compared the performance of these models on the training and validation sets using metrics such as ROC-AUC, accuracy, precision, recall, and F1 score.ResultsThe combined model, incorporating YOLOv8 for object detection and secondary classification, demonstrated excellent performance in diagnosing tonsillar and adenoid hypertrophy, significantly improving diagnostic accuracy and consistency. The ResNet18 model, due to its lightweight nature and minimal computational resource requirements, performed exceptionally well in the YOLOv8-ResNet fusion model for detecting and classifying tonsils and adenoids, making it our preferred model.ConclusionThe deep learning model combining YOLOv8n and ResNet18 based on lateral nasopharyngeal X-rays demonstrates significant advantages in diagnosing pediatric tonsillar and adenoid hypertrophy.https://www.frontiersin.org/articles/10.3389/fonc.2025.1508525/fulltonsillaradenoidartificial intelligence in medicineResNet18YOLOv8diagnostic imaging
spellingShingle Zhiqing Wu
Ran Zhuo
Yali Yang
Xiaobo Liu
Bin Wu
Jian Wang
Optimized deep learning model for diagnosing tonsil and adenoid hypertrophy through X-rays
Frontiers in Oncology
tonsillar
adenoid
artificial intelligence in medicine
ResNet18
YOLOv8
diagnostic imaging
title Optimized deep learning model for diagnosing tonsil and adenoid hypertrophy through X-rays
title_full Optimized deep learning model for diagnosing tonsil and adenoid hypertrophy through X-rays
title_fullStr Optimized deep learning model for diagnosing tonsil and adenoid hypertrophy through X-rays
title_full_unstemmed Optimized deep learning model for diagnosing tonsil and adenoid hypertrophy through X-rays
title_short Optimized deep learning model for diagnosing tonsil and adenoid hypertrophy through X-rays
title_sort optimized deep learning model for diagnosing tonsil and adenoid hypertrophy through x rays
topic tonsillar
adenoid
artificial intelligence in medicine
ResNet18
YOLOv8
diagnostic imaging
url https://www.frontiersin.org/articles/10.3389/fonc.2025.1508525/full
work_keys_str_mv AT zhiqingwu optimizeddeeplearningmodelfordiagnosingtonsilandadenoidhypertrophythroughxrays
AT ranzhuo optimizeddeeplearningmodelfordiagnosingtonsilandadenoidhypertrophythroughxrays
AT yaliyang optimizeddeeplearningmodelfordiagnosingtonsilandadenoidhypertrophythroughxrays
AT xiaoboliu optimizeddeeplearningmodelfordiagnosingtonsilandadenoidhypertrophythroughxrays
AT binwu optimizeddeeplearningmodelfordiagnosingtonsilandadenoidhypertrophythroughxrays
AT jianwang optimizeddeeplearningmodelfordiagnosingtonsilandadenoidhypertrophythroughxrays